Discourse and the Non-Native English Speaker By Michael Cribb

of coherence through, for example, the calculation of type-token ratios (e.g., Wigglesworth, 1997). In the present study, the depth and breadth of the detour into fluency will necessarily be limited in order that it does not in any way distract from the main consideration regarding coherence.

According to Gatbonton and Segalowitz (1988), fluency for non-native speakers is important since it appears to improve self-confidence, which allows students to take more risks with the language. This, in turn, allows them to participate in more conversation (Kirkland, 1984), leading to more practice and, hence, greater fluency. Fluency also allows students to circumlocute around linguistic gaps in their own knowledge, a useful communication strategy, enabling them to negotiate for meaning with native speakers. In addition, the increase in confidence and participation in conversation that comes with fluency can lead to greater motivation and pleasure in language learning (Gatbonton & Segalowitz, 1988). But what does it actually mean to say that a speaker is fluent? Certainly we cannot just equate fluency with rate of delivery of syllables, since this would qualify students who merely repeated themselves as fluent. Hedge (1993, p. 275) has suggested that fluency is the “ability to link units of speech together with facility and without strain or…undue hesitation”, while non-fluent speech is “discernible in frequent pauses, repetitions, and self-corrections”. Example 2.5 from my own field notes gives an extreme example of disfluency in non-native speech.

Schmidt (1992) suggested that fluent speech is “automatic, not requiring much attention or effort…”, while non-fluent speech is “effortful…exhibit[ing] many hesitations and other manifestations of groping for